In recent years, guilt has gradually become one of the hotspots of psychology as a self-conscious emotion. As a barometer of morality, the developmental characteristics of guilt have great effect upon the individual’s mental health. In this research, the developmental characteristics of junior high school students’ guilt and the influence of guilt on their fair behavior were investigated. First of all, the guilt scenario stories were collected and verified from two groups which were randomly selected from junior high school students (90 in group 1, 107 in group 2) as the measure materials for developmental characteristics of guilt in the following study. The materials included two kinds of scenario (family vs. peer) and each scenario contained three stories. Following each story, there were five questions which intended to find out the developmental characteristics of junior high school students’ guilt. After a series of tests, we found that the measure materials’ content validity was 0.81 and its internal consistency coefficient was 0.86. All of them indicated that the questionnaire of guilt had good reliability and validity and it could be used as measure tool of the following study. Then in study 1, the guilt scenario stories which had been collected and verified were used to investigate the developmental characteristics of guilt of 335 junior high school children from 7-9 grades. The results were as follows: The score of junior high school children’s guilt showed a gradually decreasing trend with the growth of grade, M=4.43, SD=.62; M=4.37, SD=.59; M=4.30, SD=.51. The score of guilt of girls was significantly higher than boys, F(1,317)=14.37, p=.000＜.001, η2=.04. The score of guilt without judgment was significantly higher than that with the judgment of teachers. The score of guilt in the family scenario was significantly higher than that in the peer scenario. Both the score of guilt without judgment and with peer judgment in the family scenario were significantly higher than that in the peer scenario. In study 2, we used dictator game to explore fair behavior of 120 junior high school children (evenly divided between boys and girls), with situation and emotional state as independent variables. The results were as follows: The score of guilt of the emotional arousal group was significantly higher than that of the control group, t(1,118)=-9.86, p=.000＜.001, 1-β=.99. It meant that the guilt emotional arousal method was effective. In the guilt arousal group, the number of tokens that junior high school students assigned to the partner was more than that to themselves in two person situation, but there was not significant difference in the control group. The number of tokens that they assigned to the partner, to themselves and to the third person was decreasing in turn in three person situation, but there was not significant difference from the control group. The research shows that junior high school students’ guilt can promote their fair behavior in two person situation, but it has not significant influence on their fair behavior in three person situation.